cingulate cortex (ACC), whereas craving correlated positively with activity in these regions. It is interesting to note that most fMRI studies evaluating the effects of stimulant drugs on reward report a deactivation of BOLD signals in the NAc rather than BOLD increases as may have been expected by activation of this brain region. In contrast, increases in BOLD are associated with craving (25). The deactivation of the NAc by cocaine is consistent with studies in nonhuman primates that also show inhibition of the NAc with cocaine administration, which were interpreted to reflect the predominance of D2R (inhibit cyclic AMP) over D1R (increase cyclic AMP) in the NAc (26). One important aspect to consider in fMRI studies assessing the effects of acute cocaine is that for ethical reasons these cannot be done in non-drug-abusing controls. Thus, it is not possible to determine if the BOLD responses in the NAc reflect the pharmacological effects of cocaine or, alternatively, the adaptations associated with repeated drug use that in turn alter subsequent responses to the drug. This is because stimuli (including drugs) that induce fast and large DA increases also trigger conditioned responses, as a result of which the conditioned stimuli can increase DA
